Home

Welcome to the OmniPress MCP Wiki 🌐

The Universal Agentic Publishing Protocol & Distribution Engine for AI Agents via Model Context Protocol (MCP)

OmniPress MCP is an open-source, domain-agnostic publishing protocol designed to give AI agents the power to transform raw thoughts, codebases, research papers, court rulings, policy memos, or market data into authoritative long-form publications and multi-channel social campaigns.

---

🎯 The "Truly Omni" Philosophy

OmniPress is built around three core dimensions of universality:

1. Omni-Domain: It is not limited to tech or legal work. Through Modular Niche Skills, OmniPress can write with the exact tone, jargon, and depth of any profession—from software architecture to tax forensics, public governance, biomedical research, or financial markets.
2. Omni-Platform: A single core thought is automatically syndicated into formats optimized for LinkedIn, X (Twitter), Meta Threads, Reddit (targeted subreddits), Webhooks (n8n, Make, Slack), and Markdown Archives (Substack, Medium, Jusbrasil).
3. Omni-Runtime: Runs natively on Node.js (18+) with zero external dependencies as well as on Python (3.10+) FastMCP.

---

🧭 Wiki Navigation

Explore the documentation:

• Architecture & Protocol: How OmniPress handles JSON-RPC 2.0, stdio transport, and publisher adapters.
• Creating Niche Skills: Step-by-step guide to building custom skills for your specific industry.
• Supported Platforms & Adapters: Complete configuration for LinkedIn, X, Threads, Reddit, and Webhooks.
• Privacy & De-Identification: How OmniPress shields sensitive credentials, PII, and court secrecy.
• Configuration & Environment Variables: Complete reference for .env and account credentials.
• Antigravity Integration: How to register and use OmniPress inside Google Antigravity.

---

⚡ Quick Example

In Antigravity or your AI agent:

OmniPress: Analyze our newly implemented queue retry mechanism in src/workers/queue.ts.
Generate:
1. An in-depth technical article with Markdown frontmatter.
2. A leadership lessons-learned post for LinkedIn.
3. An engaging conversational post for Meta Threads.
4. An authentic problem-solution breakdown for Reddit r/webdev.

OmniPress will formulate all 4 assets, run an automatic privacy scan to mask credentials, and queue the drafts directly to your platforms upon confirmation.